Abstract Imagine being able to precisely manipulate the behavior of disease-vectoring mosquitoes to drive them away from humans and into lethal traps. Such manipulative power would revolutionize the control of devastating diseases like malaria, dengue, and Zika. Mosquito-borne disease kills millions of people every year, causes sickness in hundreds of millions more, and can have terrible long-lasting side effects, including infant microcephaly. Here I propose a bold new approach to the design of attractive and repellent odor blends for the manipulation of mosquito behavior. I will first develop two key technologies in mosquitoes ? in vivo calcium brain imaging and behavioral analysis of tethered insects. I will then use these technologies to map the neural basis of host attraction in Aedes aegypti ? the primary mosquito vector of dengue, Zika, yellow fever, and chikungunya. The mapping approach takes advantage of natural variation between divergent mosquito subspecies and is analogous to QTL mapping in genetics. However, instead of mapping behavior to genes, we will map it to patterns of neural activation in the primary olfactory center of the insect brain. Once we understand exactly which patterns of activation drive or inhibit attraction, we can then design blends that evoke these patterns and thus mimic their behavioral effects. The results have the potential to precipitate a major leap forward in vector control as well as a breakthrough in our basic understanding of odor coding and preference.